Volume 11, Number 3, 2008
© Mary Ann Liebert, Inc.
Caloric Restriction But Not Exercise-Induced Reductions
in Fat Mass Decrease Plasma Triiodothyronine
Concentrations: A Randomized Controlled Trial
Edward P. Weiss,1,3Dennis T. Villareal,1Susan B. Racette,1Karen Steger-May,2
Bhartur N. Premachandra,4Samuel Klein,1and Luigi Fontana1,5
Caloric restriction (CR) decreases circulating triiodothyronine (T3) concentration. However, it is not known if
this effect is due to body fat mass reductions or due to CR, per se. The purpose of this study was to test the
hypothesis that plasma T3concentration decreases with CR-induced reductions in fat mass but not in response
to similar decreases in fat mass that are induced by exercise. Sedentary, nonobese 50- to 60-year-old men and
women with no clinical evidence of cardiovascular or metabolic disease and not taking thyroid medications
were randomly assigned to 12 months of caloric restriction (n ? 18) or exercise-induced weight loss (n ? 17) or
to a control group (n ? 9). Body weight and composition and plasma concentrations of the thyroid hormones
T3, thyrotropin (TSH), thyroxine (T4), and free thyroxine (FT4) were measured at baseline and 12 months. Fat
mass changed significantly in the CR (?6.3 ? 1.0 kg) and exercise (?5.5 ? 1.0 kg) groups but not in the con-
trol group (?0.6 ? 1.4 kg). The changes were not significantly different between the CR and exercise groups.
Plasma T3concentration decreased in the CR group (?9.8 ? 2.0 ng/dL, p ? 0.0001) but not in the exercise
(?3.8 ? 2.1 ng/dL, p ? 0.07) or control (?1.3 ? 2.8 ng/dL, p ? 0.65) groups. TSH, T4, and FT4did not change
in any of the study groups. Twelve months of CR decreased circulating T3concentrations in middle-aged adults.
This effect does not appear to be attributable to changes in body fat mass because a comparable decrease in T3
concentration was not observed in response to an exercise-induced fat mass reduction.
metabolic rate, and may be regulated by energy balance. It
is well documented that circulating concentrations of the
thyroid hormone, triiodothyronine (T3), decrease in re-
sponse to body weight and fat mass losses induced by
caloric restriction (CR)1–4and by CR plus endurance exer-
cise training.5–8Recently, we found that serum T3concen-
trations were lower in lean persons practicing long-term
CR, than in age-, gender-, and body fat-matched endurance
runners eating a high-calorie diet.9However, the cross-sec-
tional nature of that study makes it impossible to truly de-
termine if CR, but not exercise, has an effect on circulating
closely linked to energy metabolism; they help regulate
HORMONES are important because they are
The purpose of the present study was to conduct a ran-
domized controlled trial to assess the hypothesis that body
weight and fat mass reductions induced by long-term CR,
but not those induced by increasing exercise energy expen-
diture without changing energy intake, cause a decrease in
circulating T3 concentrations in nonobese humans. This
study is part of the CALERIE trial (Comprehensive Assess-
ment of Long-term Effects of Reducing Intake of Energy;
Clinical Trials.gov identifier: NCT00099138); reports on other
outcomes have been published previously.10–13
Subjects and Methods
Sedentary, nonsmoking, 50- to 60-year-old men and post-
menopausal women with body mass index (BMI) values of
1Division of Geriatrics and Nutritional Sciences, Department of Internal Medicine, 2Division of Biostatistics, Washington University
School of Medicine, St. Louis, Missouri.
3Department of Nutrition and Dietetics, Saint Louis University, St. Louis, Missouri.
4Thyroid Specialty Laboratory, St. Louis, Missouri.
5Division of Food Science, Human Nutrition and Health, Istituto Superiore di Sanitá, Rome, Italy.
23.5–29.9 kg/m2were recruited for the study. Medical his-
tory and physical examination were used to identify and ex-
clude volunteers with cardiovascular disease, diabetes, lung
disease, uncontrolled hypertension, and evidence of malig-
nancy. Informed written consent was obtained from all par-
ticipants and the study was approved by the Human Stud-
ies Committee at Washington University School of Medicine.
A total of 48 participants started the study interventions,
with sample sizes of 19, 19, and 10 for the CR, exercise (EX),
and healthy lifestyle control (HL) groups, respectively. Data
from 1 CR group participant and 1 EX group participant were
excluded because the subjects dropped out and were not avail-
able for final testing. Data from two other participants (one
each from the EX and HL groups) were excluded because the
participants were on thyroid medications. Therefore, the sta-
tistical analyses for the present report included 18 subjects in
the CR group, 17 in the EX group, and 9 in the HL group.
Eligible participants were randomly assigned to CR, EX,
or HL group in a 2:2:1 ratio. Individualized diet and exer-
cise prescriptions for the CR and EX groups, respectively,
were calculated from baseline total energy expenditure
which was determined by the doubly labeled water method
as described previously.10The CR and EX interventions were
designed to result in the same energy deficit. The goal of the
CR intervention was to decrease energy intake by 16% from
baseline during the first 3 months and by 20% from baseline
during the remaining 9 months. To achieve a decrease in en-
ergy intake, participants were encouraged to substitute foods
with low energy density for those with high energy density
and to reduce portion sizes.
Participants randomized to the exercise intervention were
instructed to maintain energy intake at baseline levels and
to exercise in order to increase total energy expenditure by
16% for the first 3 months and by 20% for the subsequent 9
months. Participants were given exercise energy expenditure
prescriptions on a weekly basis and exercised in our facility
or on their own. Adherence to the energy expenditure goals
was assessed by using heart rate monitors that estimate ex-
ercise energy expenditure (S610, Polar Electro Oy, Kempele,
Finland). The most commonly used exercise modes were jog-
ging, walking, elliptical machine exercise, cycling, and/or
rowing on an ergometer.
Participants in the HL group were offered advice about
consuming a healthful diet and were given free access to
community-based yoga classes. Participation in dietary con-
sultations and yoga classes was rare.
Additional details about the interventions and adherence
to the interventions have been presented previously.10,12
Body weight and fat mass
Body weight was measured in the morning, after the par-
ticipants fasted overnight and while they were wearing only
a hospital gown and underwear. Fat mass was measured by
dual energy x-ray absorptiometry.10
Carbohydrate intake was measured by using 7-day food
diaries and computerized nutrition analysis (Nutrition Data
System for Research, versions, 4.05, 4.06, and 5.0, Nutrition
Coordination Center, University of Minnesota, Minneapolis,
MN). Carbohydrate consumption and several other
macronutrient and micronutrient intakes have been reported
previously.10,11,13Carbohydrate intake was included in the
present report to help in the interpretation of the results, as
carbohydrate deprivation may affect circulating thyroid hor-
Venous blood from a forearm or antecubital vein was col-
lected into tubes containing sodium ethylenediaminete-
traacetic acid (EDTA). Within 2 hours, plasma was isolated
by centrifugation (4°C and 1800g for 20 minutes) and stored
at ? ?20°C. All samples were acquired between 7:00 AM and
11:00 AM; fasting was not required. For 48 hours or more
prior to the blood draw, the participants were advised re-
frain from all exercise or vigorous physical activity. Plasma
from venous blood was analyzed for thyroxine (T4) concen-
tration by fluorescence polarization immunoassay and for
concentrations of T3, thyrotropin (TSH), and free thyroxine
(FT4) by microparticle enzyme immunoassay (Abbott Labo-
ratories, North Chicago, IL).
Analysis of covariance (ANCOVA) was used to compare
baseline to 12-month changes in outcome variables among
study groups with adjustment for baseline values. Tukey’s
tests were used for paired comparisons of mean changes.
Within-group changes were assessed using the least squares
means from the ANCOVA. Analyses were performed using
SAS software for Linux (version 9.1.3, SAS Institute Inc.,
Cary, NC). Statistical tests were two-tailed, and significance
was accepted at p ? 0.05.
The relative distribution of men and women participants
was similar across groups; CR (39% men), EX (35% men),
and HL (44% men). At baseline, the mean (? standard de-
viation [SD]) age was 55 ? 3, 59 ? 3, and 55 ? 2 years and
body mass index (BMI) was 27.1 ? 2.5, 27.0 ? 1.9, and 28.0 ?
1.4 kg/m2in the CR, EX, and HL groups, respectively.
Body weight and fat mass
Body weight and fat mass data for a slightly larger sam-
ple have been previously presented in detail10and are pre-
sented here to assist in the interpretation of the thyroid hor-
mone data. Body weight and fat mass decreased significantly
in the CR and EX groups but not in the HL group (Table 1).
The weight and fat mass changes were not different between
the CR and EX groups.
Total daily carbohydrate intake decreased in the CR group
but not in the EX or HL groups; the change in the CR group
was significantly different from that in the EX group (Table
1). In contrast, dietary carbohydrate as a percentage of en-
ergy intake remained unchanged in all groups.
WEISS ET AL. 606
Plasma T3concentration decreased in the CR group but
not in the EX or HL groups; the change in the CR group was
significantly different from that in the HL group (Table 2).
Plasma concentrations of TSH, T4, and FT4did not change
in any group (Table 2).
The results from the present study demonstrate that CR-
induced reductions in body weight and fat mass cause a sig-
nificant decrease in circulating T3concentration while simi-
lar reductions in body weight and fat mass induced by
exercise have little or no effect on plasma T3concentration.
These findings have important implications because T3
might mediate the effect of CR on oxidative stress and
longevity.4,15However, from a weight loss and weight main-
tenance perspective, they also suggest that CR-induced
weight loss may be more difficult to maintain than weight
loss induced by exercise because low T3in euthyroid indi-
viduals is predictive of long-term weight gain,16possibly be-
cause of its effect on metabolic rate. Although on average,
T3concentrations remained in the normal range, the 11% (9.8
ng/dL) decrease in the CR group would be expected to re-
sult in a 14 kcal/d decrease in resting metabolic rate (inde-
pendent of changes in lean mass) based on published equa-
tions.17To put the magnitude of this metabolic alteration in
perspective, a positive energy balance of 14 kcal/d would
result in an annual weight gain of 0.7 kg, which is compa-
rable to the typical 0.3–1.0 kg/yr adulthood weight gain in
the United States.18,19
TRIIODOTHYRONINE, CALORIC RESTRICTION, AND EXERCISE607
TABLE 1.BODY WEIGHT, FAT MASS, AND CARBOHYDRATE INTAKE
(n ? 18)(n ? 17)(n ? 9)
Fat mass (kg)a
Carbohydrate intake (g/d)
Carbohydrate intake (% energy intake)
78.9 ? 2.2
?8.3 ? 1.1b,c
76.9 ? 2.6
?6.8 ? 1.2b,c
82.9 ? 3.9
?1.3 ? 1.60.004
26.4 ? 1.3
?6.3 ? 1.0b,c
25.6 ? 1.4
?5.5 ? 1.0b,c
26.7 ? 1.2
?0.6 ? 1.40.004
243 ? 22
?29 ? 10b,d
237 ? 12
11 ? 10
289 ? 22
?6 ? 150.03
46.7 ? 2.7
2.0 ? 1.3
46.2 ? 1.5
0.4 ? 1.4
52.1 ? 2.7
?2.1 ? 2.00.25
aBody weight and fat mass data have been presented previously for a slightly larger sample13and are presented here to assist in the inter-
pretation of the thyroid hormone data.
Baseline data are arithmetic means ? SE; change data are least squares means ? SE after adjustment for baseline values.
bp ? 0.05 within group using adjusted change from ANCOVA.
cp ? 0.05 versus HL group.
dp ? 0.05 versus EX.
CR, caloric restriction; EX, exercise; HL, healthy lifestyle; SE, standard error.
TABLE 2.PLASMA CONCENTRATIONS OF THYROID HORMONES
(n ? 18)(n ? 17)(n ? 9)
1.01 ? 0.10
0.12 ? 0.1
1.31 ? 0.17
0.11 ? 0.1
1.11 ? 0.19
?0.01 ? 0.20.77
88.9 ? 2.4
?9.8 ? 2.0a,b
95.9 ? 3.3
?3.8 ? 2.1
92.6 ? 3.9
?1.3 ? 2.80.03
6.29 ? 0.23
?0.09 ? 0.2
6.66 ? 0.35
?0.35 ? 0.2
6.56 ? 0.32
0.32 ? 0.20.10
1.08 ? 0.05
?0.01 ? 0.04
0.98 ? 0.03
?0.03 ? 0.04
0.92 ? 0.05
?0.04 ? 0.060.93
ap ? 0.05 within group using adjusted change from ANCOVA.
bp ? 0.05 versus HL group.
Baseline data are arithmetic means ? SE; change data are least squares means ? SE after adjustment for baseline values.
TSH, thyroid stimulating hormone; T3, triiodotyronine; T4, thyroxine; FT4, free thyroxine; CR, caloric restriction; EX, exercise; HL, healthy
The year-long energy deficits induced by the CR and ex-
ercise interventions were similar, as evidenced by similar
changes in fat mass (Table 1). Furthermore, the decrease in
serum leptin concentration, as reported previously,11was
not different between the CR (?61%) and exercise groups
(?62%), suggesting that metabolically, the energy deficits
were sensed as similar by the adipose tissue. These results
suggest that the decrease in plasma T3concentration that has
been previously associated with reductions in body weight
and fat mass is induced by a reduction in energy intake and
not due to negative energy balance itself.
Furthermore, because body weight was stable during the
last 3 weeks of the study (3 week changes: CR, ?0.1 kg; EX,
?0.6 kg as reported in Weiss et al.12), these changes can not
be attributed to a negative energy balance at the time of fol-
The thyroid gland secretes small amounts of T3. However,
most circulating T3is produced by deiodination of T4in the
liver and other tissues.20Because TSH, T4, and FT4were not
altered by CR, our results are consistent with the notion that
T3concentrations decreased as a result of lower peripheral
conversion of T4to T3instead of a decrease in secretion from
the thyroid gland. While the precise mechanism involved in
reduced peripheral conversion of T3to T4in the CR group
is not apparent, it is possible that the reduction in ATP
and/or cofactors in the CR group may have partially inhib-
ited the type 1 T4deiodinase activity, resulting in decreased
Low circulating T3concentrations can also be caused by
hypothyroidism. However, CR did not cause clinical or sub-
clinical hypothyroidism in our participants because TSH did
not increase,21despite the reduction in circulating T3con-
centrations. In addition, low T3concentrations are associated
with elevated systemic inflammation in a condition known
as “sick euthyroid syndrome.”22However, our subjects did
not have evidence of systemic inflammation12and CR is
known to decrease markers of systemic inflammation.9,23
A limitation of the present study is that we did not measure
free T3or thyroxine-binding globulin (TBG) concentrations and
therefore do not know if the biologically active fraction of cir-
culating T3was affected by CR. However, previous human
studies suggest that total T3and free T3decrease together in
response to CR.9,24,25In response to unintentional CR during
the Biosphere 2 experiment, total and free T3concentrations
both decreased while TBG concentrations remained un-
changed.24,25Additionally, both total and free T3concentra-
tions are lower in long-term practitioners of CR, as compared
to age-matched lean endurance athletes and sedentary con-
trols.9Therefore, it seems reasonable to expect that the reduc-
tions in total T3concentration seen in the present study were
accompanied by decreases in free T3concentrations.
We specifically studied 50- to 60-year-old men and women
because one of the objectives of the larger CALERIE trial was
to determine if CR reverses the biologic effects of aging
(younger subjects would not have any aging effects to re-
verse). Although our study results may not be directly gen-
eralizable to younger individuals, previous studies have
shown that CR decreases total T3concentrations in men and
women aged 26–50 years,8,26albeit with shorter interven-
tions. Furthermore, CR has been shown to lower T3concen-
trations in both young (approximately 13 years) and old (ap-
proximately 26 years) rhesus monkeys (median and maximal
lifespan of approximately 25 and 40 years, respectively) and
the effect may be somewhat greater in younger monkeys.4
Taken together, the findings from these previous studies sug-
gest that our results are likely applicable to younger men
and women than we studied, although it is also possible that
larger T3reductions occur in young adults. In summary, a
CR-induced fat mass reduction caused a significant decrease
in plasma T3concentration. In contrast, a similar change in
fat mass induced by exercise had little or no effect on T3, al-
though the exercise effect may have been nonsignificant be-
cause of the small sample size and limited statistical power.
Studies on rodents have demonstrated that CR decreases cir-
culating T3concentration, body temperature, resting oxygen
consumption, and oxygen free radical-induced tissue dam-
age, and slows aging. The present results add to the accu-
mulating body of evidence27that humans undergo some of
the same adaptations to CR as do rats and mice.
We are grateful to the study participants for their cooper-
ation and to the staff of the Applied Physiology Laboratory
and Nurses of the General Clinical Research Center at Wash-
ington University Medical School for their skilled assistance.
This work was supported by National Institutes of Health
(NIH) Cooperative Agreement AG20487, NIH General Clin-
ical Research Center RR00036, NIH Clinical Nutrition Re-
search Unit DK56341, and the Narveen Medical Research
Foundation. Dr. Weiss was supported by NIH AG00078.
1. Herlihy JT, Stacy C, Bertrand HA. Long-term food restric-
tion depresses serum thyroid hormone concentrations in the
rat. Mech Ageing Dev 1990;53:9–16.
2. Danforth E Jr. Effects of fasting and altered nutrition on thy-
roid hormone metabolism in man. In: Hannemann G (ed.)
Thyroid Hormone Metabolism. New York: Marcel Dekker,
1986, pp. 335–358.
3. Maglich JM, Watson J, McMillen PJ, Goodwin B, Willson
TM, Moore JT. The nuclear receptor CAR is a regulator of
thyroid hormone metabolism during caloric restriction. J
Biol Chem 2004;279:19832–19838.
4. Roth GS, Handy AM, Mattison JA, Tilmont EM, Ingram DK,
Lane MA. Effects of dietary caloric restriction and aging on
thyroid hormones of rhesus monkeys. Horm Metab Res
5. Loucks AB, Callister R. Induction and prevention of low-T3
syndrome in exercising women. Am J Physiol 1993;264:
6. Loucks AB, Heath EM. Induction of low-T3 syndrome in ex-
ercising women occurs at a threshold of energy availability.
Am J Physiol 1994;266:R817–R823.
7. Katzeff HL, Selgrad C. Maintenance of thyroid hormone pro-
duction during exercise-induced weight loss. Am J Physiol
8. Heilbronn LK, de Jonge L, Frisard MI, DeLany JP, Larson-
Meyer DE, Rood J, Nguyen T, Martin CK, Volaufova J, Most
MM, Greenway FL, Smith SR, Deutsch WA, Williamson DA,
Ravussin E. Effect of 6-month calorie restriction on bio-
markers of longevity, metabolic adaptation, and oxidative
stress in overweight individuals: a randomized controlled
trial. JAMA 2006;295:1539–1548.
9. Fontana L, Klein S, Holloszy JO, Premachandra BN. Effect
WEISS ET AL.608
of long-term calorie restriction with adequate protein and Download full-text
micronutrients on thyroid hormones. J Clin Endocrinol
10. Racette SB, Weiss EP, Villareal DT, Arif H, Steger-May K,
Schechtman KB, Fontana L, Klein S, Holloszy JO, The Wash-
ington University School of Medicine CALERIE Group. One
year of caloric restriction in humans: feasibility and effects
on body composition and abdominal adipose tissue. J
Gerontol A Biol Sci Med Sci 2006;61:943–950.
11. Villareal DT, Fontana L, Weiss EP, Racette SB, Steger-May
K, Schechtman KB, Klein S, Holloszy JO. Bone mineral den-
sity response to caloric restriction-induced weight loss or ex-
ercise-induced weight loss: a randomized controlled trial.
Arch Intern Med 2006;166:2502–2510.
12. Weiss EP, Racette SB, Villareal DT, Fontana L, Steger-May
K, Schechtman KB, Klein S, Holloszy JO. Improvements in
glucose tolerance and insulin action induced by increasing
energy expenditure or decreasing energy intake: a random-
ized controlled trial. Am J Clin Nutr 2006;84:1033–1042.
13. Weiss EP, Racette SB, Villareal DT, Fontana L, Steger-May
K, Schechtman KB, Klein S, Ehsani AA, Holloszy JO. Lower
extremity muscle size and strength and aerobic capacity de-
crease with caloric restriction but not with exercise-induced
weight loss. J Appl Physiol 2007;102(2):634–640.
14. Bisschop PH, Sauerwein HP, Endert E, Romijn JA. Isocaloric
carbohydrate deprivation induces protein catabolism de-
spite a low T3-syndrome in healthy men. Clin Endocrinol
15. Tapia G, Fernandez V, Varela P, Cornejo P, Guerrero J,
Videla LA. Thyroid hormone-induced oxidative stress trig-
gers nuclear factor-kappaB activation and cytokine gene ex-
pression in rat liver. Free Radic Biol Med 2003;35:257–265.
16. Ortega E, Pannacciulli N, Bogardus C, Krakoff J. Plasma con-
centrations of free triiodothyronine predict weight change
in euthyroid persons. Am J Clin Nutr 2007;85:440–445.
17. Astrup A, Buemann B, Christensen NJ, Madsen J, Gluud C,
Bennett P, Svenstrup B. The contribution of body composi-
tion, substrates, and hormones to the variability in energy
expenditure and substrate utilization in premenopausal
women. J Clin Endocrinol Metab 1992;74:279–286.
18. Lewis CE, Jacobs DR, Jr., McCreath H, Kiefe CI, Schreiner
PJ, Smith DE, Williams OD. Weight gain continues in the
1990s: 10-year trends in weight and overweight from the
CARDIA study. Coronary Artery Risk Development in
Young Adults. Am J Epidemiol 2000;151:1172–1181.
19. Quatromoni PA, Pencina M, Cobain MR, Jacques PF,
D’Agostino RB. Dietary quality predicts adult weight gain:
findings from the Framingham Offspring Study. Obesity
(Silver Spring) 2006;14:1383–1391.
20. Tietz Textbook of Clinical Chemistry. Philadelphia: W.B.
Saunders Co., 1999.
21. Ladenson PW, Singer PA, Ain KB, Bagchi N, Bigos ST, Levy
EG, Smith SA, Daniels GH and Cohen HD. American Thy-
roid Association guidelines for detection of thyroid dys-
function. Arch Intern Med 160: 1573-1575, 2000.
22. Papanicolaou DA. Euthyroid Sick Syndrome and the role of
cytokines. Rev Endocr Metab Disord 1: 43-48, 2000.
23. Chung HY, Kim HJ, Kim JW and Yu BP. The inflammation
hypothesis of aging: molecular modulation by calorie re-
striction. Ann N Y Acad Sci 928: 327-335, 2001.
24. Weyer C, Walford RL, Harper IT, Milner M, MacCallum T,
Tataranni PA and Ravussin E. Energy metabolism after 2 y
of energy restriction: the biosphere 2 experiment. Am J Clin
Nutr 72: 946-953, 2000.
25. Walford RL, Mock D, Verdery R and MacCallum T. Calorie
restriction in biosphere 2: alterations in physiologic, hema-
tologic, hormonal, and biochemical parameters in humans
restricted for a 2-year period. J Gerontol A Biol Sci Med Sci
26. Velthuis-te Wierik EJ, Westerterp KR, van den BH. Impact
of a moderately energy-restricted diet on energy metabolism
and body composition in non-obese men. Int J Obes Relat
Metab Disord 1995;19:318–324.
27. Fontana L, Klein S. Aging, adiposity and calorie restriction.
Address reprint requests to:
Edward P. Weiss, Ph.D.
Department of Nutrition and Dietetics
3437 Caroline Street
Saint Louis University
St. Louis, MO 63104
Received: September 28, 2007
Accepted: January 2, 2008
TRIIODOTHYRONINE, CALORIC RESTRICTION, AND EXERCISE609